Recent Scientific Accomplishments

Since 2001, Kelejian has published twelve papers, all of them related to either theoretical issues in spatial econometrics, or applications thereof. Of these papers, ten are theoretical in nature, while two are applied. One, coauthored with G. Tavlas and G. Hondroyiannis is forthcoming in Open Economies Review; it relates to contagion issues in foreign exchange markets. Another paper, coauthored with D. Robinson and published in Contemporary Issues in Urban and Regional Economies (2005), relates to the study of demographic trends in US cities, 1970-1990. With regards to the papers on spatial econometrics, four of these papers are published or are forthcoming in the Journal of Econometrics. Of the other six papers, three are in regional journals while the other three are in conference volumes. In the paper "HAC Estimation in a Spatial Framework,?coauthored with Ingmar Prucha, they suggest a non-parametric estimator of an asymptotic variance covariance matrix which would naturally arise in a spatial framework in which an IV procedure is used to estimate the model parameters. They show consistency, and give conditions under which their estimator would be positive-semidefinite in small samples. In the paper, "Panel Data Models with Spatially Correlated Error Components" with M. Kapoor and I. Prucha, they consider a spatial model in a panel data framework and extend the usual error component specification so that it includes spatial correlation. They suggest estimators and formally demonstrate, among other things, their consistency. Another paper, coauthored with Ingmar Prucha is "Estimation of Simultaneous Systems of Spatially Interrelated Cross Sectional Equations.?In this paper the authors formally specify a system of simultaneous cross sectional equations in which there are spill-overs between the dependent variables, independent variables and in the error terms. Kelejian and Prucha propose estimators for the parameters of this system and formally prove their consistency and asymptotic normally. The paper, "On the Asymptotic Distribution of the Moran I Test Statistic with Applications,?coauthored with Ingmar Prucha, gives a new central limit theorem, which accounts for triangular arrays, and uses it to derive the large sample distribution of the test statistic for spatial correlation that is typically used in spatial models, namely the Moran I statistic.

Funded Research

Kelejian is currently co-investigator (with David Drukker and Ingmar Prucha, P.I.s) on an SBIR with Stata Corp. funded by the National Institute of Aging on “New Methods and Software for Spatial-Regression Analysis.?The aims of the SBIR are to (1) develop new estimation methods for static and dynamic panel data models with spatial (cross sectional) interactions; and (2) to implement these methods in Stata to make them conveniently accessible to a wide group of researchers in the social sciences. Kelejian and Prucha were also jointly PIs of a National Science Foundation grant on the “Specification and Estimation of Spatial Models,?2000 - 2003.

Future Plans

Kelejian plans on continuing his work in spatial econometrics, both in theoretical and applied applications.